Summary

Abstract

As the Web becomes the predominant environment for more and more people to create applications and export information, syntactic approaches for navigation and keyword-based searches are becoming increasingly inadequate. In this book we have presented our proposal for enhancing query processing in a global information system. The following are some of the features that we have considered relevant:

• Providing semantic descriptions of data repositories using ontologies. This enables the user to view the repositories at the level of its relevant semantic concepts. Thus an information request can now be expressed using these concepts, and the user can now browse multiple domain ontologies as opposed to individual heterogeneous repositories or concepts based on statistical information.

• Dealing with different vocabularies so that users are not forced to use a common one. Moreover, the interontology relationships that virtually link the different ontologies (which allow the management of different vocabularies) are by themselves useful metadata about the different semantics coexisting in the global information system.

• Definition of a strategy that permits the incremental enrichment of answers by visiting new ontologies. Users are free to pose their queries using their favorite ontology, and the system is in charge of finding and retrieving the relevant data; the semantic and syntactic heterogeneity is hidden to users.

• Combination of partial translations to keep the original semantics of the user query. Thus, although only partial translations were achieved, the retrieved data will satisfy all the constraints in the user query. With this technique, the more ontologies the system visits, the more chances the system has to retrieve all the relevant data existing under each visited ontology.

• Management of imprecise answers and estimation of the loss of information incurred. This process should be limited by the user who specifies the maximum loss allowed. The system explores the different possibilities and chooses the plan with the least loss to enrich the answer.